(738h) A Conformational Analysis of an Engineered Laminin-Mimetic, Elastin-like Fusion Protein Using Molecular Dynamics Simulations

Introduction: In the central nervous system, laminin is a vital component of the extracellular matrix that is responsible for a number of cellular functions such as adhesion, migration, survival, and differentiation, and is involved in early embryonic basement membrane assembly. The fifth globular domain located at the C-terminus of the laminin α2 chain (LG5) module plays a major role as a heparin and α-dystroglycan (α-DG) binding site¹. Heparin is a highly anionic, sulfated glycosaminoglycan that helps bind exogenous growth factors that regulate and maintain neural stem cell differentiation². In neural cells, the α-DG glycoprotein complex plays a fundamental role in facilitating new laminin polymerization at the cell surface and supporting cellular adhesion³. LG5 also has a biologically active site associated with the binding to integrin β1, which is the most abundant and multifunctional integrin subunit of laminin receptors. The LG5 domain provides the functional core of our simulated protein that regulates cell-protein-matrix interactions, whereas the self-assembling structural module of the protein is composed of a designer elastin-like polypeptide (ELP) sequence. ELPs undergo a thermally-triggered inverse phase transition where they phase separate at high temperatures⁴. This reversible lower critical solution temperature behavior can be imparted to soluble proteins fused to an ELP⁵. Due to their thermal responsive behavior, recombinant ELP fusion proteins can be easily purified using inverse transition cycling⁶, vastly reducing the need for expensive chromatographic resins and improving scalability for large-scale production. In order to expand on new opportunities of exploiting the self-assembling properties of these ELPs as a route to developing novel biomaterials, a more meticulous investigation into the assembly process, as well as the interfacial properties of these proteins are needed.

Methods: We adopted the crystal structure1DYK⁷ for the LG5 domain. The ELP[K₂I₂L₂K₂]₁ sequence was built with the program Avogadro⁸ using the peptide builder tool. We assumed an alpha-helix starting structure with Φ = -60° and Ψ= -40° angles. Atomistic MD simulations were performed using NAMD 2.10⁹ and the CHARMM36 force field for proteins. Temperatures were maintained using a Langevin thermostat. The production times for the simulations were 100 ns.

Results and Discussion: The first discrete event in protein folding within the elastin-like polypeptide (ELP) domain is the loss of the α-helix secondary structure and a transition to β type secondary structures. At temperatures below 305 K, we observed a collapse of the ELP domain with respect to its initial starting conformation. At 310 K, there is a noticeable transient unfolding of the ELP domain starting at ~75 ns, highlighted by the loss of intra-strand hydrogen bonds. The sudden appearance of intra- β-strands at 310 K also suggests an inverse phase transition at that temperature. This distinct folding transition to higher ordered structures has been shown to predict self-assembly in vitro observed by other groups¹⁰. Biomaterials incorporating this self-assembling property allow them to be highly dynamic, and can self-heal following external shear forces. An analysis into the hydration properties of the ELP domain reveals that as the temperature increases, the number of surrounding water molecules decreases and the number of peptide-peptide hydrogen bonds increases. There is a slight decrease in number of surrounding water molecules over time in all simulated temperatures, however, there is an abrupt reduction of water molecules at 64 ns and 82 ns at 310 K. This change in the hydration structure is correlated with the formation of β-sheets within the trajectory.

Conclusions: We introduce a realistic design strategy that could transform current approaches to biomaterials development and address the particular challenges of characterizing the dynamic processes that occur in biological matrices. Our simulations provide both an immediate, high-resolution view of the conformational dynamics of our system, as well as a robust and extensible platform for future work directed toward our overarching goal of creating self-assembling, protein-based biomaterials as candidates for cell-delivery therapeutics.

References: [1] Urushibata, S. et al., Arch Biochem Biophys 2010, 497, (1-2), 43-54. [2] Nugent, M. A. et al., Int J Biochem Cell B 2000, 32, (2), 115-120. [3] Muntoni, F. et al., Neurotherapeutics 2008, 5, (4), 627-632. [4] Long, M. M. et al. Biophys. J. 1975, 15, (2), A71-A71. [5] Christensen, T. et al., Biomacromolecules 2013, 14, (5), 1514-1519. [6] Meyer, D. E. et al., Nat Biotechnol 1999, 17, (11), 1112-1115. [7] Hohenester, E. et al., Molecular Cell 1999, 4, (5), 783-792. [8] Hanwell, M. D. et al., J. Cheminfo. 2012, 4. [9] Phillips, J. C. et al., J Comput Chem 2005, 26, 1781. [10] Dong, H. et al., JACS 2007, 129, (41), 12468-12472.